Portal de Periódicos da CAPES

Peritoneal transport assessment by peritoneal equilibration test with 3.86% glucose: A long-term prospective evaluation

2006; Elsevier BV; Volume: 69; Issue: 5 Linguagem: Inglês

10.1038/sj.ki.5000183

1523-1755

Vincenzo La Milia, Pietro Pozzoni, Giovambattista Virga, Monica Crepaldi, Lucia Del Vecchio, Simeone Andrulli, Francesco Locatelli,

Potassium and Related Disorders

The peritoneal equilibration test (PET) with 3.86% glucose concentration (3.86%-PET) has been suggested to be more useful than the standard 2.27%-PET in peritoneal dialysis (PD), but no longitudinal data for 3.86%-PET are currently available. A total of 242 3.86%-PETs were performed in 95 incident PD patients, who underwent the first test during the first year of treatment and then once a year. The classical parameters of peritoneal transport, such as peritoneal ultrafiltration (UF), D/D0, and D/PCreat, were analyzed. In addition, the absolute dip of dialysate sodium concentration (ΔDNa), as an expression of sodium sieving, was studied. D/D0 was stable, and a progressive decrease in UF was observed after the second PET, whereas D/PCreat firstly increased and then stabilized. ΔDNa was the only parameter showing a progressive decrease over time. On univariate analysis, D/D0 and ΔDNa were found to be significantly associated with the risk of developing UF failure (risk ratio (RR) 0.987 (0.973–0.999), P=0.04, and RR 0.768 (0.624–0.933), P=0.007, respectively), but on multivariate analysis only ΔDNa showed an independent association with the risk of developing UF failure (RR 0.797 (0.649–0.965), P=0.020). UF, D/D0, and D/PCreat changed only in those patients developing UF failure, reflecting increased membrane permeability, whereas ΔDNa significantly decreased in all patients. The 3.86%-PET allows a more complete study of peritoneal membrane transport than the standard 2.27%-PET. ΔDNa shows a constant and significant reduction over time and is the only factor independently predicting the risk of developing UF failure in PD patients. The peritoneal equilibration test (PET) with 3.86% glucose concentration (3.86%-PET) has been suggested to be more useful than the standard 2.27%-PET in peritoneal dialysis (PD), but no longitudinal data for 3.86%-PET are currently available. A total of 242 3.86%-PETs were performed in 95 incident PD patients, who underwent the first test during the first year of treatment and then once a year. The classical parameters of peritoneal transport, such as peritoneal ultrafiltration (UF), D/D0, and D/PCreat, were analyzed. In addition, the absolute dip of dialysate sodium concentration (ΔDNa), as an expression of sodium sieving, was studied. D/D0 was stable, and a progressive decrease in UF was observed after the second PET, whereas D/PCreat firstly increased and then stabilized. ΔDNa was the only parameter showing a progressive decrease over time. On univariate analysis, D/D0 and ΔDNa were found to be significantly associated with the risk of developing UF failure (risk ratio (RR) 0.987 (0.973–0.999), P=0.04, and RR 0.768 (0.624–0.933), P=0.007, respectively), but on multivariate analysis only ΔDNa showed an independent association with the risk of developing UF failure (RR 0.797 (0.649–0.965), P=0.020). UF, D/D0, and D/PCreat changed only in those patients developing UF failure, reflecting increased membrane permeability, whereas ΔDNa significantly decreased in all patients. The 3.86%-PET allows a more complete study of peritoneal membrane transport than the standard 2.27%-PET. ΔDNa shows a constant and significant reduction over time and is the only factor independently predicting the risk of developing UF failure in PD patients. The peritoneal equilibration test (PET) was introduced into clinical practice by Twardowski et al.1.Twardowski Z.J. Nolph K.D. Khanna R. et al.Peritoneal equilibration test.Perit Dial Bull. 1987; 7: 138-147Google Scholar to investigate the transport characteristics of the peritoneal membrane and to give data to tailor the dialysis prescription in patients on peritoneal dialysis (PD). The original method is performed using a single peritoneal dwell 4-h long and a solution with a glucose concentration of 2.27% (2.27%-PET). During the test, the plasma and dialysate creatinine concentrations and the dialysate glucose concentration at the start and at the end of the test are assessed; peritoneal ultrafiltration (UF) is also measured. According to the transport characteristics with respect to the dialysate-to-plasma ratio of creatinine (D/PCreat) and the dialysate glucose concentration at the end of the test compared to the start (D/D0), PD patients are categorized as low (L), low-average (L-A), high-average (H-A), and high (H) transporters. UF capacity is also classified by the same method.1.Twardowski Z.J. Nolph K.D. Khanna R. et al.Peritoneal equilibration test.Perit Dial Bull. 1987; 7: 138-147Google Scholar Peritoneal UF failure is an important cause of reduced patient and technique survival in PD patients.2.Jager K.J. Merkus M.P. Dekker F.W. et al.Mortality and technique failure in patients starting chronic peritoneal dialysis: results of the Netherlands Cooperative Study on the Adequacy of Dialysis. NECOSAD Study Group.Kidney Int. 1999; 55: 1476-1485Abstract Full Text Full Text PDF PubMed Scopus (159) Google Scholar, 3.Ates K. Nergizoglu G. Keven K. et al.Effect of fluid and sodium removal on mortality in peritoneal dialysis patients.Kidney Int. 2001; 60: 767-776Abstract Full Text Full Text PDF PubMed Scopus (323) Google Scholar, 4.Brown E.A. Davies S.J. Rutherford P. et al.Survival of functionally anuric patients on automated peritoneal dialysis: The European APD outcome study.J Am Soc Nephrol. 2003; 14: 2948-2957Crossref PubMed Scopus (318) Google Scholar The PET with 3.86% glucose dialysate concentration (3.86%-PET) could be more useful than the standard 2.27%-PET to evaluate peritoneal UF in PD patients,5.Mujais S. Nolph K.D. Gokal R. et al.Evaluation and management of ultrafiltration problems in peritoneal dialysis. International Society for Peritoneal Dialysis Ad Hoc Committee on Ultrafiltration Management in Peritoneal Dialysis.Perit Dial Int. 2000; 20: S5-S21PubMed Google Scholar as the larger drained volume reduces the likelihood of measurement errors. In addition, the 3.86%-PET is capable of giving an estimate of aquaporin-1-mediated water transport;6.Rippe B. Stelin G. Simulations of peritoneal solute transport during CAPD. Application of two-pore formalism.Kidney Int. 1989; 35: 1234-1244Abstract Full Text PDF PubMed Scopus (186) Google Scholar, 7.Krediet R.T. Lindholm B. Rippe B. Pathophysiology of peritoneal membrane failure.Perit Dial Int. 2000; 20: S22-S42PubMed Google Scholar the reduction in dialysate sodium concentration usually observed during a hypertonic dwell, a phenomenon called sodium sieving (SNa), is explained by transcellular free-water moving through endothelial channels impermeable to solutes (aquaporin-1), which therefore causes dilution of the dialysate over the first 2 h of the dwell.8.Stelin G. Rippe B. A phenomenological interpretation of the variation in dialysate volume with dwell time in CAPD.Kidney Int. 1990; 38: 465-472Abstract Full Text PDF PubMed Scopus (88) Google Scholar, 9.Rippe B. Stelin G. Haraldsson B. Computer simulations of peritoneal fluid transport in CAPD.Kidney Int. 1991; 40: 315-325Abstract Full Text PDF PubMed Scopus (244) Google Scholar, 10.Carlsson O. Nielsen S. Zakaria E.L.-R. Rippe B. In vivo inhibition of transcellular water channels (aquaporin-1) during acute peritoneal dialysis in rats.Am J Physiol. 1996; 271: H2254-H2262PubMed Google Scholar At the same time, a number of studies have shown that the glucose concentration does not influence the peritoneal transport of small solutes and the related indexes, such as D/PCreat,11.Virga G. Amici G. da Rin G. et al.Comparison of fast peritoneal equilibration tests with 1.36 and 3.86% dialysis solutions.Blood Purif. 1994; 12: 113-120Crossref PubMed Scopus (23) Google Scholar, 12.Smit W. Langedijk M.J. Schouten N. et al.A comparison between 1.36 and 3.86% glucose dialysis solution for the assessment of peritoneal membrane function.Perit Dial Int. 2000; 20: 734-741PubMed Google Scholar, 13.Pride E.T. Gustafson J. Graham A. et al.Comparison of a 2.5 and a 4.25% dextrose peritoneal equilibration test.Perit Dial Int. 2002; 22: 365-370PubMed Google Scholar and reference values for solute and fluid transport using a modified 3.86%-PET have been recently evaluated in a large group of prevalent PD patients.14.Smit W. van Dijk P. Langedijk M.J. et al.Peritoneal function and assessment of reference values using a 3.86% glucose solution.Perit Dial Int. 2003; 23: 440-449PubMed Google Scholar However, unlike the standard 2.27%-PET, for which time changes of peritoneal UF and small solutes transport characteristics are available,15.Davies S.J. Phillips L. Giffiths A.M. et al.What really happens to people on long-term peritoneal dialysis?.Kidney Int. 1998; 54: 2207-2217Abstract Full Text Full Text PDF PubMed Scopus (305) Google Scholar, 16.Heimbürger O. Wang T. Lindholm B. Alterations in water and solute transport with time on peritoneal dialysis.Perit Dial Int. 1999; 19: S83-S90PubMed Google Scholar, 17.Wong T.Y.H. Szeto C.C. Lai K.B. et al.Longitudinal study of peritoneal membrane function in continuous ambulatory peritoneal dialysis: relationship with peritonitis and fibrosing factors.Perit Dial Int. 2000; 20: 679-685PubMed Google Scholar, 18.Davies S.J. Phillips L. Naish P.F. Gavin I.R. Peritoneal glucose exposure and changes in membrane solute transport with time on peritoneal dialysis.J Am Soc Nephrol. 2001; 12: 1046-1051PubMed Google Scholar, 19.Davies S.J. Longitudinal relationship between solute transport and ultrafiltration capacity in peritoneal dialysis patients.Kidney Int. 2004; 66: 2437-2445Abstract Full Text Full Text PDF PubMed Scopus (167) Google Scholar no longitudinal data have so far been collected with the 3.86%-PET. This study was aimed at evaluating the variations over time of the peritoneal transport of small solutes, of SNa and of UF in a cohort of incident PD patients, using the 3.86%-PET. Ninety-five patients (M/F 45/50), who underwent at least one 3.86%-PET within 12 months from the start of PD, were examined and followed up for a median of 25 months (total range 4–153 months) (Figure 1). A total of 242 3.86%-PETs were performed, without any complications (except only one patient experiencing cramps). Primary diagnoses of renal failure were glomerulonephritis 33 (34.7%), hypertensive nephropathy 14 (14.7%), adult polycystic renal disease 12 (12.6%), reflux nephropathy/interstitial nephritis 14 (14.7%), diabetic nephropathy 11 (11.6%), and small kidneys/unknown 11 (11.6%). At the start of PD, patients had a median age of 61 years (total range 27–83 years); the median duration of PD at the time of the first PET was 3.6 months (inter-quartile range 2.9–5.0 months). Baseline peritoneal transport characteristics, obtained during the first 3.86%-PET, were as follows: UF 737±228 ml, D/D0 0.23±0.06, D/PCreat 0.71±0.09, ΔDNa 9.7±3.0 mmol/l. According to these baseline values, patients were categorized into four transporter groups (Table 1). A number of baseline factors significantly correlated, albeit weakly, with UF. On multivariate analysis, gender, age at start of PD therapy, urine volume, D/PCreat, and ΔDNa remained significant independent covariates, accounting for 25% of the variability (Table 2). In 13 patients with two repeated (after 48 h) 3.86%-PETs, the pooled coefficient of variation for the net UF volume was found to be 7.8%.Table 1Number of patients assigned to different transporters categories, as assessed by baseline D/PCreat, D/D0, and UF values (n=95)D/PCreat (%)D/D0 (%)an=94.UF (%)High17 (17.9)16 (17.0)12 (12.6)High-average33 (34.7)35 (37.2)38 (40.0)Low-average33 (34.7)27 (28.7)28 (29.5)Low12 (12.6)16 (17.0)17 (17.9)D/PCreat, dialysate-to-plasma ratio of creatinine concentration at the end of the PET; D/D0, ratio of dialysate glucose concentration at the end and at the start of the PET; UF, peritoneal ultrafiltration.a n=94. Open table in a new tab Table 2Multivariate regression of baseline factors associated with baseline UFStandardized β coefficienttP-valueModel constant5.06<0.001Sex (female)-47.4-2.10.043Age at start of PD (1 year)-3.3-2.10.036Urine volume (1 ml)-0.07-2.40.018D/PCreat (0.001)-0.60-2.280.025ΔDNa (1 mmol/l)17.32.250.027ΔDNa, absolute dip of sodium dialysate sodium concentration at 60 min. See Table 1 for other abbreviations.ANOVA for model P<0.001, R2=0.25. Open table in a new tab D/PCreat, dialysate-to-plasma ratio of creatinine concentration at the end of the PET; D/D0, ratio of dialysate glucose concentration at the end and at the start of the PET; UF, peritoneal ultrafiltration. ΔDNa, absolute dip of sodium dialysate sodium concentration at 60 min. See Table 1 for other abbreviations. ANOVA for model P<0.001, R2=0.25. Patients were then evaluated according to the minimal number of PETs performed, and independently analyzed as different cohorts, as described above (Figure 2). In 62 patients, at least two PETs were performed; in these patients, UF and D/D0 remained stable, D/PCreat increased and ΔDNa decreased between the first and the second PET. In 42 patients in whom at least three PETs were performed, UF showed a significant reduction at the third PET, D/D0 was substantially stable, D/PCreat increased at the second PET but showed a reduction at the third PET with a general trend that was not statistically significant, and ΔDNa significantly decreased over time. In 20 patients in whom at least four PETs were performed, UF showed a significant reduction at the third PET, D/D0 and D/PCreat remained stable, and ΔDNa significantly decreased over all the PETs. Finally, in 11 patients in whom at least five PETs were performed, the only variable that showed a significant variation over time was ΔDNa, which halved in 5 years; also UF progressively decreased, but the reduction was not statistically significant. When analyzing the mean differences of peritoneal transport characteristics between the last and the first PET in all patients with at least two PETs performed, UF and ΔDNa were the only parameters showing a significant reduction (Table 3). The reduction of UF and ΔDNa were –3.0±10.2 ml/month of PD therapy and -0.09±0.12 mmol/l/month of PD therapy, respectively.Table 3Mean differences in peritoneal transport characteristics between the last and the first PET in patients who performed at least two 3.86%-PETs (n=62)Mean differencePUF (ml)-126±3340.004D/D00.00 (-0.04–0.04)0.601D/PCreat0.02±0.110.127ΔDNa (mmol/l)-3.1±4.0<0.001Data are expressed as means±s.d. or as median with inter-quartile range within brackets.See Tables 1 and 2 for abbreviations. Open table in a new tab Data are expressed as means±s.d. or as median with inter-quartile range within brackets. See Tables 1 and 2 for abbreviations. A number of factors correlated, albeit weakly, with UF decrease over time. However, on multivariate analysis, excluding baseline UF because of autocorrelation, no factor was predictive of UF decrease (data not shown). During the study, 15 out of 95 patients developed peritoneal UF failure, as defined above (two patients at the first, three patients at the second, four patients at the third, three patients at the fourth, and three patients at the fifth PET). The proportion of patients without UF failure was 98% (95% confidence intervals: 94–100%) at 1 year, 92% (95% confidence intervals: 85–99%) at 2 years, and 84% (95% confidence intervals: 73–95%) at 3 years (Figure 3). With the Cox univariate regression model, only D/D0 and ΔDNa were found to be significantly associated with the risk of developing UF failure (risk ratio (RR) 0.987 (0.973–0.999), P=0.04, and RR 0.768 (0.624–0.933), P=0.007, respectively) (Table 4). However, on Cox multivariate analysis, only ΔDNa remained significantly associated with the risk of developing UF failure (RR 0.797 (0.649–0.965), P=0.020).Table 4Cox univariate regression model for the development of UF failure during the follow-up in 95 patients on PD therapyRisk ratioLower CIUpper CIPSex (female)0.7540.4381.3360.369Age at start of PD (increase of 1 year)1.0100.9701.0530.618Urine volume (increase of 1 ml)0.9990.9991.0000.995GFR (increase of 1 ml/min)0.9060.6961.1340.406Plasma Albumin (increase of 1 g/dl)0.4640.1141.7420.259UF (increase of 1 ml)0.9980.9971.0000.225D/D0 (increase of 0.001)0.9870.9730.9990.043D/PCreat (increase of 0.001)1.0050.9981.0120.161ΔDNa (increase of 1 mmol/l)0.7680.6240.9330.008GFR, glomerular filtration rate; CI, confidence interval; see Tables 1 and 2 for other abbreviations. Open table in a new tab GFR, glomerular filtration rate; CI, confidence interval; see Tables 1 and 2 for other abbreviations. No differences were observed between patients with and without UF failure as to baseline characteristics (Table 5), whereas the number of peritonitis and the number of patients with at least one episode of peritonitis during the follow-up were higher in the group with UF failure, although not significantly (0 (0–1) vs 1 (0–2), P=0.250, and 9 (69%) vs 21 (45%), P=0.209, respectively).Table 5Baseline characteristics of patients without and with UF failure who performed at least two 3.86%-PETs (n=60)UF NormalUFFPAge (years)58±1255±180.504M/F23/245/80.500Plasma Albumin (g/dl)3.87±0.393.84±0.420.853Urine volume (ml)1000 (600–1400)520 (235–1295)0.202GFR (ml/min)2.6 (1.5–4.4)1.6 (0.4–3.2)0.148UF (ml)744±243771±1090.697D/D00.23±0.050.21±0.040.346D/PCreat0.72±0.090.73±0.050.603ΔDNa (mmol/l)10.0±3.48.4±3.00.103GFR, glomerular filtration rate; see Tables 1 and 2 for other abbreviations.Data are expressed as means±s.d. or as median with inter-quartile range within parentheses. Open table in a new tab GFR, glomerular filtration rate; see Tables 1 and 2 for other abbreviations. Data are expressed as means±s.d. or as median with inter-quartile range within parentheses. Peritoneal UF and membrane transport characteristics (D/D0 and D/PCreat) remained stable in the group without UF failure, whereas patients developing UF failure showed a reduction of UF and an increase of membrane permeability (as indicated by increased D/PCreat and decreased D/D0); however, ΔDNa significantly decreased in both groups (Figure 4). The correlations of the differences between the first and the last PET plotted against the average of the initial and final values were small for all the examined parameters (UF: r=0.19; D/D0: r=0.31; D/PCreat: r=0.24; ΔDNa: r=0.23), suggesting a small influence of the phenomenon of regression to the mean. The 2.27%-PET has become the standard of care for PD prescription. More recently, the 3.86%-PET has been introduced to provide better information on UF, because the larger drained volume reduces the likelihood of measurement errors, and on free-water transport by means of SNa. Indeed, the coefficient of variation of UF was only 7.8% in a subgroup of patients of our study, whereas it has been reported as being closer to 50% with the use of 2.27%-PET.19.Davies S.J. Longitudinal relationship between solute transport and ultrafiltration capacity in peritoneal dialysis patients.Kidney Int. 2004; 66: 2437-2445Abstract Full Text Full Text PDF PubMed Scopus (167) Google Scholar UF, its variations, and its relationship with other parameters of solute transport through the peritoneal membrane are therefore better assessed with the use of the 3.86%-PET. Furthermore, the 3.86%-PET is also capable of giving an estimate of aquaporin-1-mediated free-water transport. In our knowledge, this study represents the first prospective evaluation of peritoneal transport properties assessed by 3.86%-PET in a large group of incident PD patients. Its findings could be of particular relevance when considering the advice of the ISPD committee on UF failure, who recommended the use of the 3.86%-PET in everyday clinical practice instead of the 2.27%-PET.5.Mujais S. Nolph K.D. Gokal R. et al.Evaluation and management of ultrafiltration problems in peritoneal dialysis. International Society for Peritoneal Dialysis Ad Hoc Committee on Ultrafiltration Management in Peritoneal Dialysis.Perit Dial Int. 2000; 20: S5-S21PubMed Google Scholar Indeed, reference values for parameters of peritoneal solute transport obtained with this method are still lacking in incident PD patients. In this study, we provided the baseline characteristics of peritoneal solute transport of all incident patients with at least one 3.86%-PET performed within 12 months from the start of PD. In general, we found higher transport of low molecular weight solutes compared to those obtained in a similar study by Smit et al.14.Smit W. van Dijk P. Langedijk M.J. et al.Peritoneal function and assessment of reference values using a 3.86% glucose solution.Perit Dial Int. 2003; 23: 440-449PubMed Google Scholar In particular, in our study, D/PCreat and UF were higher and D/D0 was lower than those found in that recent study.14.Smit W. van Dijk P. Langedijk M.J. et al.Peritoneal function and assessment of reference values using a 3.86% glucose solution.Perit Dial Int. 2003; 23: 440-449PubMed Google Scholar These differences could be explained by the fact that Smit et al.14.Smit W. van Dijk P. Langedijk M.J. et al.Peritoneal function and assessment of reference values using a 3.86% glucose solution.Perit Dial Int. 2003; 23: 440-449PubMed Google Scholar calculated their reference values in patients who had been on PD therapy over a wide range of time (3 months–12 years) after excluding patients with UF failure (approach A), or analyzing all patients within their first 2 years of PD therapy without the exclusion of patients with UF failure (approach B). Our baseline data are more homogenous, given that the first PET was always performed in the first year from the start of treatment. We also prospectively examined peritoneal SNa, as ΔDNa at 60 min of the test, peritoneal transport of low molecular solutes and UF in all PD patients with more than one 3.86%-PET performed. The analyses were performed separately in the different cohorts obtained by dividing the patients according to the minimal number of PETs performed, because the systematic loss of patients occurring throughout the follow-up could have distorted the real mean values of the single parameters. For this reason, we decided to compare the baseline properties of peritoneal solute transport at the first PET with those at subsequent PETs in the same cohort of patients. An interesting finding of this study is that UF and ΔDNa, and particularly the latter, were the only two parameters showing a significant variation over time in all the groups. On the contrary, the classical parameters of peritoneal transport remained substantially stable throughout the follow-up. In particular, D/PCreat showed an increase, especially at the second PET, but over time this increase was not statistically significant, whereas D/D0 remained stable. This is in agreement with the findings of previous studies,20.Rocco M.V. Jordan J.R. Burkart J.M. Changes in peritoneal transport during the first month of peritoneal dialysis.Perit Dial Int. 1995; 15: 12-17PubMed Google Scholar, 21.Johnson D.W. Mudge D.W. Bizzard S. et al.A comparison of peritoneal equilibration tests performed 1 and 4 weeks after PD commencement.Perit Dial Int. 2004; 24: 460-465PubMed Google Scholar indicating that, after an initial increase of small solute transport, the majority of patients treated with continuous ambulatory peritoneal dialysis for up to 3 years experience a stability of peritoneal UF capacity and of small-solute transport characteristics. The small correlation of the changes between the first and the last PET plotted against the average of the initial and final value for all parameters suggest a small influence of the phenomenon of regression to the mean on the longitudinal changes of membrane characteristics.22.Altman D.G. Practical Statistics for Medical Research. Chapman & Hall, London1991: 284-285Google Scholar, 23.Blake P.G. Abraham G. Sombolos K. et al.Changes in peritoneal membrane transport rates in patients on long term CAPD.Adv Perit Dial. 1989; 5: 3-7PubMed Google Scholar However, the main finding of this study is the different evolution over time of membrane characteristics in patients with or without UF failure. First of all, among baseline parameters, only ΔDNa independently predicted the development of UF failure. Having elevated ΔDNa at baseline seems to be protective against the subsequent development of UF failure; this parameter could therefore be useful to identify those patients who are at a higher risk of developing UF failure already at the time of starting PD. In this study, we could not detect any factor being responsible for the different baseline ΔDNa levels observed in the studied PD patients. However, we did not take into consideration potentially relevant factors, such as, for example, inflammatory markers, and at the same time, we did not perform any genetic evaluation in our patients. Another important finding of this study is that whereas membrane permeability, expressed as UF, D/D0, and D/PCreat, remained stable in patients without UF failure, and increased (UF and D/D0 decreased and D/PCreat increased) in those developing UF failure, parameters related to SNa, as ΔDNa, decreased in both groups of patients. It has been suggested by computer simulations that the reduction of ΔDNa could be the result of reduced UF occurring over time.24.Rippe B. de Arteaga J. Venturoli D. Aquaporins are unlikely to be affected in marked ultrafiltration failure: results from a computer simulation.Perit Dial Int. 2001; 21: S30-S34PubMed Google Scholar However, in our study the reduction of ΔDNa always preceded that of peritoneal UF and was observed even in patients with stable UF. These results therefore suggest that the first alteration occurring in the peritoneal membrane characteristics of PD patients might be a reduction of the osmotic conductance to glucose (LpSσg) through aquaporin-1 channels, resulting in decreased ΔDNa, possibly because of the chronic exposure to peritoneal fluid.25.Rippe B. Venturoli D. Simonsen O. de Arteaga J. Fluid and electrolyte transport across the peritoneal membrane during CAPD according to the three-pore model.Perit Dial Int. 2004; 24: 10-27PubMed Google Scholar However, only in a subgroup of patients was this alteration associated with an increase of peritoneal permeability to small solutes, and only in these patients UF failure was observed, therefore suggesting that an increased peritoneal permeability is a crucial requisite for the development of UF failure. It has been hypothesized that the increased transport of small solutes through the peritoneal membrane could be caused by neoangiogenesis occurring in the peritoneum, leading to an increase in the effective peritoneal surface area.26.Mateijsen M.A. van der Wal A.C. Hendriks P.M. et al.Vascular and interstitial changes in the peritoneum of CAPD patients with peritoneal sclerosis.Perit Dial Int. 1999; 19: 517-525PubMed Google Scholar, 27.Combet S. Miyata T. Moulin P. et al.Vascular proliferation and enhanced expression of endothelial nitric oxide synthase in human peritoneum exposed to long-term peritoneal dialysis.J Am Soc Nephrol. 2000; 11: 717-728PubMed Google Scholar The increased transport of small solutes observed in patients with UF failure could also at least partly explain why the reduction of ΔDNa seemed to occur at a higher rate in these patients, as it also implies an increased diffusive transport of sodium from plasma to dialysate, finally resulting in a more consistent reduction of ΔDNa. It is possible to study the SNa of the peritoneal membrane during a hypertonic dwell through parameters other than ΔDNa, such as the ratio of D/P sodium concentration (D/PNa) at 60 min or the variation of D/PNa between the start and at 60 min of the test (ΔD/PNa). Nonetheless, in our opinion ΔDNa is the most simple and reliable tool to asses free-water transport, because D/PNa and ΔD/PNa are also dependent on the variation of plasma sodium concentration over time and on the method used to measure plasma sodium concentration.28.La Milia V. Di Filippo S. Crepaldi M. et al.Spurious estimations of sodium removal during CAPD when [Na]+ is measured by Na electrode methodology.Kidney Int. 2000; 58: 2194-2199Abstract Full Text Full Text PDF PubMed Google Scholar, 29.La Milia V. Di Filippo S. Crepaldi M. et al.Sodium removal and sodium concentration during peritoneal dialysis: effects of three methods of sodium measurement.Nephrol Dial Transplant. 2004; 19: 1849-1855Crossref PubMed Scopus (21) Google Scholar To our knowledge, this is the first report showing the variation of UF assessed by the 3.86%-PET over a long period of PD therapy. Furthermore, we showed some interesting alterations of the function of peritoneal membrane over time, such as the progressive decrease of the absolute dip of dialysate sodium concentration, expression of aquaporin-1 function, a stable permeability of the membrane in patients without UF failure, and an increased permeability of the membrane in patients with UF failure. Based on the results of this study, it may be argued that the 3.86%-PET could substitute the classical 2.27%-PET, as it allows a better exploration of the function of the peritoneal membrane. ΔDNa appears as the first and most suitable marker of peritoneal membrane function as to UF capacity. If these results will be confirmed by other studies, ΔDNa could then become the main parameter to be considered when studying early and long-term modifications of the peritoneal membrane. The same parameter could also be very useful when comparing the effect of different dialysis solutions and procedures on the preservation of UF and of the transport characteristics of the peritoneal membrane. After the publication of the theory of the ‘three peritoneal pores’,6.Rippe B. Stelin G. Simulations of peritoneal solute transport during CAPD. Application of two-pore formalism.Kidney Int. 1989; 35: 1234-1244Abstract Full Text PDF PubMed Scopus (186) Google Scholar, 9.Rippe B. Stelin G. Haraldsson B. Computer simulations of peritoneal fluid transport in CAPD.Kidney Int. 1991; 40: 315-325Abstract Full Text PDF PubMed Scopus (244) Google Scholar, 30.Rippe B. A three-pore model of peritoneal transport.Perit Dial Int. 1993; 13: S35-S38PubMed Google Scholar we began to use the 3.86%-PET to study the peritoneal UF and the transport characteristics of the peritoneal membrane. Ninety-five consecutive incident patients starting PD at the A Manzoni Hospital, Lecco, Italy, were studied from January 1993 to June 2005, after having given their informed consent. In all these patients, a 3.86%-PET was performed during the first 12 months from the start of PD, and then once a year. All patients were treated with lactate-buffered, conventional dialysis solutions. At the time of the test, all the patients had been peritonitis-free for at least 4 weeks. The dwell before the PET (overnight dwell, from 1100 to 0700 hours) was performed using a 2 l PD solution containing a glucose concentration of 1.36%, with lactate as the buffer. Blood samples were drawn at the start of the 3.86%-PET, and fresh PD fluid (Dt0') samples were taken from the bag at the end of the infusion. After the complete infusion of a 2 l 3.86% glucose PD solution, 20 ml dialysate samples were taken at 1, 60, 120, and 240 min (Dt1', Dt60', Dt120', Dt240') after flushing back 30 ml of dialysate. The patients were instructed to sit up or move about in bed before the drawing of each dialysate sample; otherwise, they remained recumbent during the test. After 240 min, the dialysate was collected by gravity for at least 20 min. The volume of the infused fresh PD solution and the drained dialysate was measured by weighing the bag and then subtracting the weight of the empty bag; no corrections were made for differences in the specific weight of the solutions. UF failure was defined as a net UF of less than 400 ml at the end of a 4-h dwell with 3.86% glucose solution, according to the definition proposed by the International Society for Peritoneal Dialysis Committee on UF failure.5.Mujais S. Nolph K.D. Gokal R. et al.Evaluation and management of ultrafiltration problems in peritoneal dialysis. International Society for Peritoneal Dialysis Ad Hoc Committee on Ultrafiltration Management in Peritoneal Dialysis.Perit Dial Int. 2000; 20: S5-S21PubMed Google Scholar In 13 patients, the 3.86%-PET was repeated 48 h later to evaluate the coefficient of variation for the net UF volume. Plasma and dialysate creatinine, total protein, and glucose concentrations were analyzed using a Hitachi 717 (Hitachi, Ltd, Tokyo, Japan); dialysate creatinine concentration was assessed by an enzymatic method in order to eliminate the effect of the high dialysate concentration. Total dialysate sodium concentration was analyzed twice using an IL 943 flame photometer (Instrumentation Laboratory, Milan, Italy). D/D0 was calculated as the ratio of dialysate glucose concentration between the end and the start of the PET. D/PCreat was calculated as the dialysate solute concentration at the end of the PET divided by creatinine plasma concentration. Plasma water concentrations were used to calculate D/PCreat.31.Waniewski J. Heimbürger O. Werynski A. Lindholm B. Aqueous solute concentrations and evaluation of mass transport coefficients in peritoneal dialysis.Nephrol Dial Transplant. 1992; 7: 50-56PubMed Google Scholar To study SNa, the absolute dip of dialysate sodium concentration (ΔDNa) at 60 min of the PET was used, calculated as follows:ΔDNa(mmol/l)=NaDialysate(mmol/l)at the start of the PET−Na Dialysate(mmol/l)at60min Both glucose and Na dialysate concentrations at the start of the PET were measured in fresh PD fluid samples taken from the bag. Results were expressed as mean values±1 s.d. or ±1 s.e. for normally distributed data. Median values and inter-quartile range were given for asymmetrically distributed data. Mean values ±1 s.d. of D/PCreat, D/D0, and UF were used to categorize PD patients at baseline, as reported elsewhere.1.Twardowski Z.J. Nolph K.D. Khanna R. et al.Peritoneal equilibration test.Perit Dial Bull. 1987; 7: 138-147Google Scholar Patients were divided into cohorts according to the minimal number of PETs performed, so that changes of the transport characteristics of the peritoneal membrane were analyzed in the same patients. Repeated measures analysis of variance was used to evaluate differences of the same variable between the first and the following PETs within each cohort. Paired t-test and Wilcoxon's signed-rank test were used to evaluate differences of the same variable between the first and the last PET for data with normal and asymmetrical distribution, respectively. In those patients who underwent a second 3.86%-PET 48 h after the first, standard deviation of the net UF was summarized by pooling the individual (SDi) of the N individual patients (Ni):Pooled SD=√∑(SDi)2/Ni The pooled coefficient of variation for the net UF volume was then obtained from the pooled SD (coefficient of variation=pooled SD/mean). The Kaplan–Meier curve was used to evaluate the time for the development of UF failure. The prognostic value for the development of UF failure of several variables was analyzed by univariate and multivariate Cox's regression analysis. In the first step, all covariates that were significantly associated with UF failure (P<0.05 at univariate Cox regression analysis) were identified. Then, the predictive power for UF failure of all significant variables was tested on multivariate Cox's models. RRs and their 95% confidence intervals were calculated with the use of the estimated regression coefficients and their standard errors in the Cox regression analysis. Multiple analysis of variance was used to compare changes over time of the studied parameters between patients with and without UF failure. To evaluate the phenomenon of regression to the mean, changes of UF, D/PCreat, D/D0, and ΔDNa between the first and the last PET were plotted against the average of the initial and final values, as suggested by Altman.22.Altman D.G. Practical Statistics for Medical Research. Chapman & Hall, London1991: 284-285Google Scholar Pearson's correlation coefficient (r) was calculated for linear correlation analysis between D/D0, D/PCreat, ΔDNa, and UF. When more than one linear correlation was found, multiple regression was used to explore the dominant relationship. A P-value of ≤0.05 was considered as significant. All the statistical analyses were performed using JMP (SAS Institute Inc.) for Windows statistical software (release 4.0.0). The invaluable assistance of the peritoneal nursing staff of Alessandro Manzoni Hospital is gratefully acknowledged.